Software upgrade on all nodes

Long time?

new node will be prepared, customers will be moved over to the new node (downtime between 30 seconds to 10 minutes) but then we will not move VPSs back - after VPS migration from 3.0 to 3.0sp1 is completed there will be no additional downtime.

My box on vz03-ca has been out for a couple of hours since the migration started. Is there a problem with this server?
 
My box on vz03-ca has been out for a couple of hours since the migration started. Is there a problem with this server?

So far we're synchronizing data from vz03-ca to the destination system. This operation doesn't involve any downtime and/or any other problems. VPS reboot will happen only after all data will be synched over. Please submit a ticket if you have problems with your VPS.
 
The last post mentioned copying data...how will this process go if you have a database that's changing say, every minute? Will the data on the new node be in sync when switched over?
 
No data will be lost at all. Migration is done in the following way:
1. VPS data is synched to the destination server. This process takes anywhere between couple minutes and couple hours per VPS;
2. Actual VPS migration is started and is done in the following major steps:
2a) Special file modification monitoring agent is being started to intercept all files modification requests and to store a list of modified files;
2b) rsync of VPS data is started - this process goes through every single directory and file and syncs changes to the destination system. This rsync might take minutes or hours depending on number of files;
2c) VPS is stopped on the source system;
2d) final rsync is executed in order to synchronize newly created and/or updated files to the destination system. This rsync process will not go through the whole directory structure but will work only with files, collected by the monitoring agent which was started in step 2a as such this operation doesn't take much time in most cases;
2e) VPS is started on the destination system.

The above scheme guarantees that all data will be copied over and no any changes will be lost at all.
Basically we're doing initial "long" rsync in order to reduce amount of time required for actual migration - most of the files inside the VPS won't be changed between points when first rsync is executed and when actual migration is started. As such first real migration rsync won't take much time which. in turn, means that very little number of files will be updated during the pre-shutdown data synchronization. In most cases the only files that get updated are various logs (Apache, mail server, etc) and database files.

Regards,
Paul
 
All the backups saved on VZPP hae been wiped and i can no longer save backups.. is this normal? any one else experiencing this?

This is the error i get:

Operation backup with the VPS(s) VEID811 as incremental, remote to 65.99.235.231 is finished with errors: #2502 2006-12-24T08:15:46-0500 vzbackup(14760): Starting backup. Nodes - vz12.2006-12-24T08:15:46-0500 vzbackup(14760): Starting node vz12 backup..
 
Top